Completing Yo r Q alit Completing Your Quality Improvement Plan
Completing Yo r Q alit Completing Your Quality Improvement Plan Completing Yo r Q alit Completing Your Quality Improvement Plan
Statistical Variation in Measures• Random variation in all measures• Greater variability if:– Rare event– Smaller hospital– Small sample size• Can lead to difficulty in making determinationthat a true improvement took place6
Approach to Statistical Variation• Step 1: consider target for an indicator• Step 2: calculate l % improvement anticipatedi t • Step 3: examine (using simple guidelines in next slides, orformal statistical tests) whether this level of improvementcould be detected statistically• Step 4: if it cannot, consider the following options:– Consider a more ambitious target– Consider making this a multi‐year target instead; do not tie exec compto indicator this year– Consider other indicators for the same problem, with more frequentobservations or greater ability to make large change– Base the baseline on longer time period (e.g. past 2 yrs)– Collect more data (e.g. more surveys)7
- Page 1: Completing Your QualityImprovement
- Page 4 and 5: What are Required Indicators?• Fo
- Page 8 and 9: Rare EventsIf you had a bad eventTh
- Page 10 and 11: Patient Experience SurveyThis comin
- Page 12 and 13: Seasonal Variation• Some indicato
- Page 14 and 15: Rare EventsIf you had a bad eventTh
- Page 16 and 17: Ideas for Improvement:Change Sectio
- Page 18 and 19: Example• Our plan: improving the
- Page 20 and 21: Template: Line‐by‐Line StyleImp
- Page 22 and 23: “Improvement Initiative”• Opt
- Page 24 and 25: Falls• Hospitals tracking falls i
- Page 26 and 27: VAP, CLI- Follow BundlesVAP:• Hea
- Page 28 and 29: • Special considerations:C Diffic
- Page 30 and 31: ALC, ED waits, Readmissions• Thre
- Page 32 and 33: HSMR• Possible goals: reduction 5
- Page 34 and 35: Patient Experience ‐ Data• Curr
- Page 36 and 37: Setting Performance Goals and Targe
- Page 38 and 39: Part A- Overview• Provide plain l
- Page 40 and 41: Part A ‐What the organization wil
- Page 42 and 43: Part A ‐ Ideas for ImprovementFro
- Page 44 and 45: How the plan aligns with the other
Approach to Statistical Variation• Step 1: consider target for an indicator• Step 2: calculate l % improvement anticipatedi t • Step 3: examine (using simple guidelines in next slides, orformal statistical tests) whether this level of improvementcould be detected statistically• Step 4: if it cannot, consider the following options:– Consider a more ambitious target– Consider making this a multi‐year target instead; do not tie exec compto indicator this year– Consider other indicators for the same problem, with more frequentobservations or greater ability to make large change– Base the baseline on longer time period (e.g. past 2 yrs)– Collect more data (e.g. more surveys)7